The Significance of Groundwater Flow Modeling Study for Simulation Of

Home , Cone of depression, Groundwater flow

water

Article The Signiﬁcance of Groundwater Flow Modeling Study for Simulation of Opencast Mine Dewatering, Flooding, and the Environmental Impact

Jacek Szczepi ´nski

Poltegor-Institute, Institute of Opencast Mining, 51-616 Wrocław, Poland; [email protected]

Received: 21 January 2019; Accepted: 18 April 2019; Published: 23 April 2019

Abstract: Simulations of open pit mines dewatering, their flooding, and environmental impact assessment are performed using groundwater flow models. They must take into consideration both regional groundwater conditions and the specificity of mine dewatering operations. This method has been used to a great extent in Polish opencast mines since the 1970s. However, the use of numerical models in mining hydrogeology has certain limitations resulting from existing uncertainties as to the assumed hydrogeological parameters and boundary conditions. They include shortcomings in the identification of hydrogeological conditions, cyclic changes of precipitation and evaporation, changes resulting from land management due to mining activity, changes in mining work schedules, and post-mining void flooding. Even though groundwater flow models used in mining hydrogeology have numerous limitations, they still provide the most comprehensive information concerning the mine dewatering and flooding processes and their influence on the environment. However, they will always require periodical verification based on new information on the actual response of the aquifer system to the mine drainage and the actual climate conditions, as well as up-to-date schedules of deposit extraction and mine closure.

Keywords: open pit; dewatering; ﬂooding; groundwater modeling; Poland

1. Introduction In the mining industry, water-related problems are among the most important aspects which can decide whether a new mine development will be reasonable or even feasible. For an assessment into the costs of mining operations, the rate of mine water inflow, dewatering technology, and the environmental impact of mine drainage are all important factors. In the past, one would mainly focus on the hazards connected with mine water inflow; at present, most of the attention is focused on flooding the post-mining voids and the environmental impact assessment [1]. At the initial stage of research, to estimate the mine water inflow, one typically applies the methods of hydrogeological analogy or hydrogeological balance. Hydrogeological calculations are usually performed using classical analytical methods described in numerous textbooks [2–5]. Further investigations are dominated by numerical methods. They are used throughout the mine’s whole life-cycle. Numerical methods allow forecasting the process of mine dewatering as well as the process of flooding the post-mining excavations with greater accuracy than any other method. They can assess the impact of dewatering on groundwater, surface water, water chemistry, water intakes, soil, flora, farmlands and forests, land subsidence, and others. Particularly, the process of flooding post-mining voids with complex geometry and diversified hydrogeological conditions must rely on numerical models [6]. Modeling methods were widely used in mining hydrogeology since the beginnings of their application. Nevertheless, the groundwater flow modeling applied in the mining industry belongs to

Water 2019, 11, 848; doi:10.3390/w11040848 www.mdpi.com/journal/water Water 2019, 11, 848 2 of 16 the most demanding methods in hydrogeology. Because mine dewatering and groundwater rebound are changing in time and in space, they should be solved in transient conditions. In the area covered by the dewatering impact, both boundary conditions and hydrogeological parameters need to be updated in response to hydrogeology and hydrology conditions as well as scope and schedule of mining works changing due to mining activities. The topics related to modeling study on mine dewatering and its environmental impact, including groundwater/surface-water interactions, groundwater resources, and hydrogeochemistry, have been the subject of many scientific publications which are mentioned in this paper. Papers on mine water and its impact on the environment appear mainly in the International Mine Water Association post-conference proceedings and in scientific journals, especially Mine Water and the Environment [7–9], Groundwater [10–12], Environmental Geology [13,14], and others [15]. In addition, the following textbooks deserve special attention: Mine Water Hydrology, Pollution, Remediation (in which mine water hydrology and chemistry are presented with an explanation of the complexities of mine water pollution and the hydrogeological context of its formation) [16], and Construction Dewatering and Groundwater Control (which presents the theory and practice of dewatering from the engineering point of view) [5]. In recent years, extensive monographs on the water management related to flooding of post-mining voids have appeared: Water Management at Abandoned Underground Flooded Mines. Fundamentals, Tracer Tests, Modeling, Water Treatment [17]; Mine Pit Lakes: Characteristics, Predictive Modeling, and Sustainability (Management Technologies for Metal Mining Influenced Water) [18]; and Acidic Pit Lakes. The Legacy of Coal and Metal Surface Mine [19]. The aim of the paper is to present the specificity of the application of groundwater flow modeling for simulation of opencast mine dewatering, flooding, and the environmental impact. This paper is a result of many years of the author’s experience on groundwater flow modeling for open-pit dewatering and its flooding as well as any environmental impact caused by mine dewatering in Poland and abroad.

2. Application of Numerical Modeling in Mining Hydrogeology The electric analogue computer model using the method of electrohydrodynamic analogy appeared in the 1930s in the oil industry in the United States. Numerical models in calculations of groundwater flow model have been used since the 1960s. In 1965, a paper was published on the use of these methods for solving the groundwater flow equation [20], and, in 1968, a digital-computer program was written to solve linear, parabolic, and partial-differential equations based upon an implicit finite-difference technique [21]. A significant increase in the use of numerical methods occurred in the first half of the 1970s, with the appearance of the Prickett-Lonnquist Aquifer Simulation Model (PLASM) [22] and the U.S. Geological Survey (USGS) model [23], which solved the three-dimensional groundwater flow model based on the finite difference method. The history of the beginnings of groundwater flow modeling was extensively presented by Bredehoeft [24]. In the mining industry, groundwater flow models are applied in a wide range using both the finite difference method and the finite element method for porous aquifers as well as fissured aquifers. Currently, both in Poland and around the world, the MODFLOW program based on the finite difference method [25,26] and the FEFLOW program based on the finite element method [27] are used the most extensively in mining hydrogeology. The modeling results usually simultaneously include a prediction of the mine water inflow, an input parameters into an environmental impact assessment, and the water management of the post-mining excavation. The present state of groundwater flow modeling applications in mining hydrogeology was presented by Rapantova et al. [9]. In mining, groundwater flow modeling represents both regional (aquifer system) and local issues (mining excavation area). Numerical modeling is used during each stage of the mining operation, beginning with deposit exploration to mine decommissioning with mine site rehabilitation. Its purpose is to deliver reliable forecasts of mine water inflow and the input parameters into an environmental impact assessment of mine drainage with reference to different scenarios of the deposit opening up and Water 2019, 11, 848 3 of 16 its extraction, as well as rehabilitation of post-mining excavation. The results of numerical simulations enable to provide reliable data which help with the decision making at each stage of mining operation:

Stage of identification, documentation, and opening up of the deposit: Assessment of • hydrogeological conditions of the deposit area, mine water inflow calculation, suggestion on dewatering technology, determination of the cone of depression and its boundaries, impact assessment on surface water (watercourses and reservoirs) and groundwater in each dewatering layer, determination of the area of environmental impact of mining, and indication the necessary measures to minimize the negative impact on the environment. Stage of deposit exploitation: Forecasting the mine water inflow intensity and the change of • groundwater level resulting in changes of geotechnical conditions of slopes in time and space as the mine develops, as well as determining impacts on the environment based on real data gathered during previous dewatering. Stage of mine decommissioning: An indication of a rational way of flooding and management the • post-mining void, determination of time span for filling an excavation with water, final water level in the post-mining reservoir, and the forecasting of quantitative and qualitative changes in the reservoir and its environment during and after flooding.

In many cases, models are also accomplished in order to conﬁrm or reject the mine drainage impact on the environment. For example, the results of numerical simulation can provide the proof on real impact of mine dewatering and/or climatic conditions on groundwater level and groundwater resources.

3. Specificity of Application the Modeling Methods for Opencast Mining The procedure for developing groundwater flow models in surface mining conditions generally complies with the overall modeling methodology which is presented in many textbooks and guidelines [28–31]. However, it must be adjusted to the specificity of open pit mine drainage operations.

3.1. Groundwater Flow System Representation Groundwater flow in the vicinity of open pits or underground mines is three-dimensional in most cases; consequently, 3-D numerical groundwater flow models must be based on 3-D hydrogeological data [32] and should be simulated by a full three-dimensional model, with the hydraulic head simulated for each model layer. It is particularly important in the areas adjacent to the open pits, where the groundwater flow has a substantial and important 3-D component. For an assessment of the range and extent of the cone of depression or the mine water inflow, one can apply a quasi-three-dimensional model which only represents the vertical flow through semi-permeable layers (Figure1)[33]. A simulation of mine drainage and its influence on the water environment needs a huge amount of data—more than for other regional models. In order to develop reliable forecasts, it is necessary to recognize the hydrogeological and hydrological conditions of the deposit and neighborhood aquifers as well as to identify all environmental, mining and technological factors which can affect the mine water inflow. The most important factors are: hydrogeological parameters, dewatering technology, the recharge of aquifers from precipitation and surface water as well as management of post-mining excavations. The conceptual model requires not only detailed the identification of hydrogeological conditions within of the model area but also a correct representation of the mine drainage system (Section 3.2). The watercourses, lakes, and reservoirs within the model area should be simulated, taking into consideration their possible drying as well as the re-wetting in the case of groundwater rebound. General principles of accepting internal and external boundary conditions have been presented extensively in many studies [28,34–37]. An important step in modeling study is the development of a conceptual model of recharge processes. It is formed by integrating many factors: Spatial and temporal variability in recharge, climate, soil and geology, surface topography, hydrology, depth to water, and others [38]. The term “potential Water 2019, 11, 848 4 of 16 recharge” was introduced by Rushton [39]. This type includes the excesses of precipitation over evapotranspiration, which subsequently disappear through a local discharge system or by evaporation from the saturated zone, but which could become a “permanent” recharge by the lowering of a shallow water table after extraction. Moreover, lowering a shallow water table can induce additional recharge by reducing evapotranspiration. This concept of potential recharge is important for modeling future conditions [40]. In Poland, the groundwater table in pre-mining condition is frequently 1–3 m below the land surface elevation, so evapotranspiration is a very important factor in groundwater balance. In mining areas, due to the extinction of evapotranspiration from groundwater as a result of a groundwater table lowering (as well as evaporation from land surface and zone of aeration), induced groundwater resources (part of a groundwater renewable resources) are created in the area of the cone of depression [41].

N A S metre a.s.l. N A C D G S 200 160 Q 120 Tn 80 40 Tw 0 Tp -40 A -80 M -120 0 500m -160

B Recharge from precipitation and rivers

Mine dewatering system Groundwater intakes

LAYER I

LAYER II Boundary inflow and outflow Boundary inflow and outflow

1 2 3 4 5 6 M 7 8 G S 9 10 11

Figure 1. Simplified geological cross-section N-S with the conceptual model suitable for numerical modelling. Explanation: 1—porous, permeable formations; 2—impermeable and slightly permeable formations; 3—fissured-karstic formations; 4—lignite; 5—boundary of the Quaternary aquifer; 6—faults; 7—symbols of aquifers; 8—open-pit limits; 9—dewatering wells barriers; 10— direction of groundwater flow; 11—direction of groundwater percolation [33].

Recharge and evaporation boundary conditions should ensure the appropriate representation of effective infiltration and its changes during water table fluctuation whenever such changes take place. Generally, recharge in groundwater models can be represented with constant head boundary conditions, specified-flux boundary conditions, mixed-boundary conditions, or some combination of the mixed and flux boundary conditions. Modeling studies where transient drawdown cones are developing at pumping centers are an example where a specified-flux boundary is appropriate for recharge estimation. Water 2019, 11, 848 5 of 16

However, for variably controlled recharge, the modeler can invoke mixed-boundary conditions or some combination of the mixed and ﬂux boundary conditions [42].

3.2. Simulation of Opencast Mine Dewatering System The opencast mine dewatering system is modeled on the basis of information of the planned location, scope, and schedule of mining works related to the deposit opening up, the mining face advancement, and the development of the internal dump. Before starting the simulation, it should be determined at what time and to what depth the groundwater level should be lowered, ensuring the eﬀective deposit dewatering and safety of slopes of the active mine or post-mining voids. For this purpose, it is necessary to obtain information from the top and bottom of the deposit seam as well as the forecasted bench levels. Guidelines for mine drainage are developed on the basis of assumptions regarding the time advance with which the overburden should be dewatered prior to the start of mining works. Under Polish law, it is assumed that the required lowering of the groundwater table (usually one meter below the bottom of the pit) should be obtained for one year before the deposit begins to be exploited. This requires gradual lowering of pressure in subsequent years. The location of the mine drainage system and the required groundwater level change, both in time and in space, must be performed in transient conditions. It can be simulated by boundary conditions of Type I: Hydraulic head as a function of time; Type II: Flux as a function of time; or Type III: Flux as a function of head. In the case of semi-permeable layers between aquifers, boundary conditions should be assigned separately for each layer on the model. The hydraulic contacts between all water-bearing layers allow for the simulation of the drainage system by the boundary condition assigned to the lowest aquifer, usually the most important from a dewatering point of view. In MODFLOW, for a mine dewatering system simulation, the Time-Variant Speciﬁed-Head Package can be used, which allows the user to specify head boundaries that can change within or between stress periods [43].

3.3. Simulation of Post Mining Voids Flooding To develop a reliable conceptual model for flooding of the post-mining void, the spatial parameters of a final excavation are required (Figure2). As a result of changing hydraulic and spatial parameters of aquifers and new stresses, which appear in the modeled area, the conceptual model used for pre-mining conditions may require improvement or even updating that represents mining or postmining conditions properly and improve the predictive numerical model. Surface water reservoirs are usually hydraulically connected to the groundwater system and can play a significant role in it. In this sense, their impact is similar to watercourses. They can discharge or recharge groundwater. Most often in groundwater models, reservoirs are simulated like rivers and represented by boundary conditions of the first or third type [45,46]. In the conditions of water filling of post-mining excavations, the water level in the reservoir in the specified time intervals is not known. Therefore, calculation methods must be used in which the water table level in the reservoir will be part of the numerical groundwater flow model solution. Such an approach is necessary, for example, when assessing changes in groundwater flow conditions in the area of reservoirs, which are under the influence of artificial drainage (mining drainage, groundwater intakes), groundwater recharge (irrigation, recharge wells, flooding), and climate change [47]. One of the ways of simulating the post-mining reservoir is the so-called method of “high conductivity cells”, which consists in assigning to the cells representing the reservoir a high value of hydraulic conductivity, usually several orders higher than the hydraulic conductivity of adjacent aquifers—for example k = 10,000 m/d [47–50]. In the case of a transient simulation, it is necessary to assign a specific yield of 1.0 [49,51]. Cells representing the reservoir on the model then become part of the solution of the numerical equation of groundwater flow, which makes it possible to calculate the surface water table level (Figures2 and3). The limitations regarding this method include, among others: 1: The method is limited to seepage lakes; 2: Changes in lake surface area are not accommodated without additional programming; 3: The Water 2019, 11, 848 6 of 16

method may require a large number of iterations to converge; 4: The magnitude of the calculated head diﬀerential across the lake should be close to zero, but it is sensitive to the imposed regional gradient and the ratio of the hydraulic conductivity of the aquifer to the hydraulic conductivity assigned to the Waterlake 2019 nodes, 11, [x47 FOR]. PEER REVIEW 6 of 17

N A S metre a.s.l. 200 160 Water reservoir Q 120 80 40 M 0 M -40 Tp -80 M -120 0 500m -160 B Recharge from precipitation and rivers

Water reservoir LAYER I

LAYER II

LAYER III Boundary inflow outflowand Boundary inflow and outflow

1234 5 6 7 M 8910 1

Figure 2. Simplified geological cross-section through the post-mining void (A) with the conceptual Figure 2. Simplified geological cross-section through the post-mining void (A) with the conceptual model suitable for numerical modeling (B). Explanation: 1—porous, permeable formations; model suitable for numerical modeling (B). Explanation: 1—porous, permeable formations; 2— 2—impermeable and slightly permeable formations; 3—fissured-karstic formations; 4—lignite; impermeable and slightly permeable formations; 3—fissured-karstic formations; 4—lignite; 5— 5—internal dump; 6— boundary of the Quaternary aquifer; 7—faults; 8—symbols of aquifers; internal dump; 6— boundary of the Quaternary aquifer; 7—faults; 8—symbols of aquifers; 9—open- 9—open-pit limits; 10—direction of groundwater flow; 11—direction of groundwater percolation [44]. pit limits; 10—direction of groundwater flow; 11—direction of groundwater percolation [44]. 3.4. Model Calibration and Verification Surface water reservoirs are usually hydraulically connected to the groundwater system and can As mine drainage is a process variable in time and space, the models used in mining hydrogeology play a significant role in it. In this sense, their impact is similar to watercourses. They can discharge are solved in a transient simulation. However, at the first stage—i.e., for natural conditions, prior to or recharge groundwater. Most often in groundwater models, reservoirs are simulated like rivers and the deposit dewatering—the model should be solved in steady-state conditions based on the average represented by boundary conditions of the first or third type [45,46]. In the conditions of water filling precipitation and evaporation data, hydrological data (flows and water levels in rivers and lakes), of post-mining excavations, the water level in the reservoir in the specified time intervals is not and hydrogeological data (groundwater levels and groundwater flows). Calibration performed under known. Therefore, calculation methods must be used in which the water table level in the reservoir steady-state conditions enables the preliminary determination of aquifer recharge as well the horizontal will be part of the numerical groundwater flow model solution. Such an approach is necessary, for and vertical hydraulic conductivity of aquifers. example, when assessing changes in groundwater flow conditions in the area of reservoirs, which are In case we have data from pumping/aquifer tests or the first period of mine dewatering, a calibration under the influence of artificial drainage (mining drainage, groundwater intakes), groundwater to transient conditions can be performed to obtain the preliminary information on the specific yield recharge (irrigation, recharge wells, flooding), and climate change [47]. and specific storage for the modeled area. In case of no such a data, the next steps of modeling One of the ways of simulating the post-mining reservoir is the so-called method of "high in transient conditions must be based on archival data. After dewatering operation starts, model conductivity cells", which consists in assigning to the cells representing the reservoir a high value of verification should proceed in transient conditions, taking into account the periods of mine dewatering. hydraulic conductivity, usually several orders higher than the hydraulic conductivity of adjacent This requires at least one-time data. This enables the model users to improve the parameter-value aquifers—for example k = 10,000 m/d [47–50]. In the case of a transient simulation, it is necessary to estimates determined in the steady-state calibration as well as the calibration to the transient condition assign a specific yield of 1.0 [49,51]. Cells representing the reservoir on the model then become part of(if the available) solution to of improve the numerical the estimated equation values of groundwater of the hydrogeological flow, which parameters makes it possible (Figure4 to). calculate the surface water table level (Figures 2, 3).

Water 2019 11 Water 2019, 11, x 848 FOR PEER REVIEW 7 of 17 16

Layer High hydraulic conductivity River 185 1 160 2 140 3 125 4 105

5 80

1 10 20 30 40 50 1

180

20 160 140 A 125

Water 2019, 11, x FOR PEER 30REVIEW 8 of 17 transient condition (if available) to improve the estimated values of the hydrogeological parameters (Figure 4). Numerical modeling40 should be understood as an iterative process in which reasonableness of results at each step should be checked before proceeding to the next step [29]. If the previously calibrated model does47 not reflect the real system reaction to the introduced stresses, the values

1 assumed in the model should be re-calibrated180 23 using different datasets. The factors in need of possible updating include the boundary conditions, aquifer and aquitards, and hydrogeological and spatial Figure 3. Simulation of a post-mining void using “high conductivity cell” method. Explanation: parameters, as well as a modification of the role of faults under the influence of deep mine drainage 1Figure—mine 3. boundarySimulation area, of a2 —surfacepost-mining contour void interval,using "hig3—meshh conductivity discretization cell" method. [33]. Explanation: 1— and leftovermine boundary rock displacements area, 2—surface due contour to mining interval, activity. 3—mesh discretization [33].

Natural conditions Hydrogeology conditions changing under mining activities TheSteady limitations state simulations regarding this method include,Transient among simulation others: 1: The method is limited to seepageMODEL lakes; CALIBRATION 2: Changes in lakeMODEL surface VERIFICATION area are not accommodatedFORECAST without additional programming; 3: The method may require a large number of iterations to converge; 4: The magnitude t 1 t 2 t 3 t 4 t 5 t 6 of the calculated head differential across the lake should be close to zero, but it is sensitive to the Model re-calibration imposed regional gradient and the ratio of the hydraulic conductivity of the aquifer to the hydraulic conductivity assigned to the lake nodes [47]. - Mine dewatering system development - Artificial sealing sections of rivers and watercouses 3.4. Model Calibration and Verification- Relocated sections of rivers, new sedimentation ponds - Change of effective infiltration due to change of groundwater level or land development As mine drainage is a- processShutdown or variable reconstruction in of groundwatertime and intakes space, the models used in mining hydrogeology are solved in -a The transient progress of simulationdeposit exploatation. However, and internal atdump the development first stage—i.e., for natural - Other phenomena (i.e.waste dumps development, recharge of aquifer by industrial water) conditions, prior to the deposit dewatering—the model should be solved in steady-state conditions basedFigure on the 4. averageCalibrationCalibration precipitation and and verification verification and evaporationof of numerical numerical data,models models hydrological in in the the opencast opencast data mining mining(flows activityand activity water area. area. levels in riversExplanation: and lakes), t1t1––t6t6: :andStress Stress hydrogeological periods; periods; t1t1: :Natural Natural data conditions conditions(groundwater (before (before levels dewatering dewatering and groundwateroperation); operation); t2t2 ––t4flows).t4:: CalibrationDewatering performed periodperiod withwith under realreal monitoring monitoringsteady-state data; data; conditionst5 t5–t6–t6: A: A forecast forecast enables of of mine minethe dewatering preliminary dewatering and and floodingdetermination flooding of the of of aquiferthepost-mining rechargepost-mining void.as wellvoid. the horizontal and vertical hydraulic conductivity of aquifers. In case we have data from pumping/aquifer tests or the first period of mine dewatering, a calibrationInNumerical the tomining transient modeling area, conditions boundary should becan conditions understood be perfor representingmed as an to iterative obtain development the process preliminary in whichof the information reasonablenessmine dewatering on the of systemspecificresults atshouldyield each and stepbe specificassigned. should storage beAdditionally, checked for the before modeled outside proceeding artheea. miningIn case to the area,of nextno itsuch stepis necessarya [data,29]. Ifthe theto next update previously steps the of boundarymodelingcalibrated in modelconditions transient does notconditionsto reflectsimulate the must realartificially be system based reactionse onaled archival or to therelocated data. introduced After rivers dewatering stresses, and watercourses, the operation values assumed starts, new constructedmodelin the modelverification ditches should should and be re-calibrated reservoirs, proceed innew using transient water different intake cond datasets.itions,s, new takingland The factorsdevelopments into account in need oftheincluding possible periods industrial updating of mine anddewatering.include municipal the boundary This waste requires dumps, conditions, at andleast aquifer a one-timechange and of aquitards,data. effective This andinfiltration enables hydrogeological the due model to a lowering andusers spatial to orimprove parameters,increase the of theparameter-value groundwater estimatestable or an determined additional in recharge the steady-sta of aquiferste calibration by industrial as well water as the from: calibration Water-based to the drilling mud used during drilling activity, damaged sewerage and water systems, and landfills (if they exist). To quantify an uncertainty in the calibrated model caused by uncertainty in the estimates of aquifer parameters, stresses, and boundary conditions, a sensitivity analysis must be performed [28]. As a result, the water balance should be presented for steady-state simulations as well as for each stage of model verification. Its correctness should be evaluated by selected calibration criteria. The results obtained during calibration and verification of the model should specify: • Hydro-structural system and aquifers transmissivity, • specific yield and storage capacity of aquifers, • changes in the leakage of water through semi-permeable layers, • recharge of aquifers and its changes due to fluctuations in the groundwater table, • sections of watercourses with decreasing or increasing flow. At the stage of a forecasted simulation, a further updating of certain boundary conditions and hydrogeological or spatial parameters of the model can be required, which should be properly assigned to represent new stresses on the model (mine dewatering system development, new intakes, ditches, and ponds) and parameters representing an open pit area with the extracted deposit, the internal overburden dump development, or the post-mining excavations.

3.5. Documenting of Modeling Study for the Mining Area

Water 2019, 11, 848 8 of 16 as well as a modification of the role of faults under the influence of deep mine drainage and leftover rock displacements due to mining activity. In the mining area, boundary conditions representing development of the mine dewatering system should be assigned. Additionally, outside the mining area, it is necessary to update the boundary conditions to simulate artificially sealed or relocated rivers and watercourses, new constructed ditches and reservoirs, new water intakes, new land developments including industrial and municipal waste dumps, and a change of effective infiltration due to a lowering or increase of the groundwater table or an additional recharge of aquifers by industrial water from: Water-based drilling mud used during drilling activity, damaged sewerage and water systems, and landfills (if they exist). To quantify an uncertainty in the calibrated model caused by uncertainty in the estimates of aquifer parameters, stresses, and boundary conditions, a sensitivity analysis must be performed [28]. As a result, the water balance should be presented for steady-state simulations as well as for each stage of model verification. Its correctness should be evaluated by selected calibration criteria. The results obtained during calibration and verification of the model should specify:

Hydro-structural system and aquifers transmissivity, • specific yield and storage capacity of aquifers, • changes in the leakage of water through semi-permeable layers, • recharge of aquifers and its changes due to fluctuations in the groundwater table, • sections of watercourses with decreasing or increasing flow. • At the stage of a forecasted simulation, a further updating of certain boundary conditions and hydrogeological or spatial parameters of the model can be required, which should be properly assigned to represent new stresses on the model (mine dewatering system development, new intakes, ditches, and ponds) and parameters representing an open pit area with the extracted deposit, the internal overburden dump development, or the post-mining excavations.

3.5. Documenting of Modeling Study for the Mining Area Documenting the results of a modeling study in mining hydrogeology is more extensive compared to models used for groundwater-resource assessment. They must account for all boundary conditions and aquifers parameters subject to changes in time. While applying a model, the natural conditions, the dewatering period, and the post-mining excavations management should be taken into consideration. The results thus obtained, including the rate of mine water inﬂow, the hydraulic head in each modeled layer, the range of the cones of depression for all layers, the groundwater balance, and the hydrogeological conditions in each stage of mining operations and after its completion—ending with post-mining area reclamation—should be presented in both descriptive and graphical form for all the periods assumed. The sensitivity analysis for the model should be presented.

4. Problem of Aquifer Drying/Rewetting and its Representation in the Model The drying/rewetting problem can be particularly damaging when undertaking groundwater modeling for mining applications [52]. In the case of confined layers where, due to dewatering, the thickness of the saturated zone does not change significantly in relation to its total thickness, the water-bearing layer may only be simulated by a transmissivity parameter. However, for a water-table aquifer and a confined layer that can be “dried” due to dewatering, it is necessary to specify on a model their top and bottom. The model may then face the problem of instability (lack of convergence) associated with the necessity of re-saturation of “dry cells” during the groundwater table rebound. In addition, due to groundwater table lowering beneath the bottom of a layer, the cells representing this layer are inactive during a calculation, and its recharge may be impossible, which is often inconsistent with the adopted conceptual model. The innovative approach was only proposed as part of the block-centered flow MODFLOW- SURFACT package (BCF4), where, taking into account all the changes introduced in previous versions, Water 2019, 11, 848 9 of 16 the pseudo-soil water retention automatically generated during the calculations was added to the formula for three-dimensional groundwater flow in a variably saturated system [53]. It does not allow for the appearance of inactive cells in areas where the water table has fallen below the bottom of the aquifer. The operation of this package consists in calculating, in each of the “dry” cells, the hydraulic height necessary to start water seepage through the unsaturated zone and not allowing the cells to be deactivated. Another solution to prevent complete drying of cells was presented by Doherty [52]. He introduced a modification to the block-centered flow package, so-called the “residual saturated thickness,” allowing the reduced water conductivity of the aquifer to be maintained even if the groundwater table falls below the base of the cell. This modification allows leaving the cell active for groundwater flow. Aquifer drying/rewetting process has a particular importance in case of deep open pit mines in which, for safety production, there are necessary dewatering many layers which are above the deposit. In a consequence of modeling study, the groundwater table on a model is lowering beneath the bottom of layers and the cells are inactive during a calculation (no groundwater flow). After the exploitation is ceased, the process of the groundwater table rebound starts, and the cells representing inactive layers haveWater 2019 to be, 11 re-activated, x FOR PEER REVIEW (re-saturated). 10 of 17 5. Use of Modeling Methods in Polish Opencast Mines Bełchatów basins—situated in the central part of Poland—and the Turów basin—situated in the south-westIn Poland, part groundwater of Poland. Generall flow numericaly, the floor models depth have of lignite been used seams since below the mid-1970s.the terrain Thesurface modeling varies studiesfrom 40–300 for simulation m. The ofthickness opencast of mine lignite dewatering occurring were in performedone to three using seams the HYDRYLIBis from 5–60 [54 m,] program, and the whichoverburden was based thickness on the is finite from di 30–200fference m. method The overburden and the FKWH constitutes [55] program, Quaternary which and was basedNeogene on theformations finite element consisting method. of silt Currently,and clays (30–75%) MODFLOW and issands the most(70–25%) popular. [60]. The The lignite numerical seams models are flat are or performedare slightly for dipping sulphur up and to ten raw degr materialsees, and opencast they are mines. sometimes However, intersected the most by extensivefaults. All groundwater deposits are flowbelow models the natural are used groundwater for the existing table, and lignite they opencast occur most mine frequently [56,57] (Figures right under5 and the6)and terrain proposed surface. ligniteAnnual opencast precipitation mines in [58 the,59 ]region (Figures of 7lignite and8). basins varies from 500–700 mm/year, and the average annual temperature is about +8 °C.

W ID

A A

W I

K B

A A

R G

G ZELÓW R AB POLAND IA

WARSZAWA A

K WROCŁAW W

IA Ó

N ILS K

N P A D

Ą I R E B A R C T BEŁCHATÓW Ó I KRAKÓW R E W

A K C A

W Z

P I K SZCZERCÓW L R I A C S A Ó ĄŻ A W N CI K WI U A DAW L KA ZB. WAWRZYKOWIZNA

A Szczerców Bełchatów L U Field Field C IĄ Ż K A N R P IE A R C S P/Bełchatów U IEC Ó D Z W KA K A

PAJĘCZNO W A ID IC AW C K P Ę A DZIAŁOSZYN IS R IA K

PRZEDBÓRZ W 8 AR RADOMSKO TA

S A MK WARTA DO RA 048 12km 1 2 3 4 5 210 6

FigureFigure 5.5.The The range range of theof numericalthe numerical model model in the areain the of thearea Bełchat of theów ligniteBełchatów mine lignite with groundwater mine with contoursgroundwater for miningcontours conditions. for mining conditions. Explanations: Explanations:1—watersheds; 1—watersheds;2—cone 2 of—cone depression of depression range; 3range;—groundwater 3—groundwater ﬂow direction; flow direction;4—hydrogeological 4—hydrogeological cross-section; cross-section;5—model boundary; 5—model6 —groundwaterboundary; 6— levelgroundwater in mining level conditions in mining [56 ].conditions [56].

N S Cone of depression range m n.p.m. Bełchatów Field

200

150

100

0 2 4km 0

5 7 89 1 2 34 6

Figure 6. Hydrogeological cross-section through Bełchatów Opencast Mine. Explanations: 1—fine sands; 2—glacial tills; 3—limestones, marls and dolomites; 4—lignite; 5—stratigraphic boundary; 6— faults; 7—screened intervals of a weeks; 8—ground water level in pre-mining conditions; 9—ground water level in mining conditions [57].

Water 2019, 11, x FOR PEER REVIEW 10 of 17

Bełchatów basins—situated in the central part of Poland—and the Turów basin—situated in the south-west part of Poland. Generally, the floor depth of lignite seams below the terrain surface varies from 40–300 m. The thickness of lignite occurring in one to three seams is from 5–60 m, and the overburden thickness is from 30–200 m. The overburden constitutes Quaternary and Neogene formations consisting of silt and clays (30–75%) and sands (70–25%) [60]. The lignite seams are flat or are slightly dipping up to ten degrees, and they are sometimes intersected by faults. All deposits are below the natural groundwater table, and they occur most frequently right under the terrain surface. Annual precipitation in the region of lignite basins varies from 500–700 mm/year, and the average annual temperature is about +8 °C.

W ID

A A

W I

K B

A A

R G

G ZELÓW R AB POLAND IA

WARSZAWA A

K WROCŁAW W

IA Ó

N ILS K

N P A D

Ą I R E B A R C T BEŁCHATÓW Ó I KRAKÓW R E W

A K C A

W Z

P I K SZCZERCÓW L R I A C S A Ó ĄŻ A W N CI K WI U A DAW L KA ZB. WAWRZYKOWIZNA

A Szczerców Bełchatów L U Field Field C IĄ Ż K A N R P IE A R C S P/Bełchatów U IEC Ó D Z W KA K A

PAJĘCZNO W A ID IC AW C K P Ę A DZIAŁOSZYN IS R IA K

PRZEDBÓRZ W 8 AR RADOMSKO TA

S A MK WARTA DO RA 048 12km 1 2 3 4 5 210 6

Figure 5. The range of the numerical model in the area of the Bełchatów lignite mine with groundwater contours for mining conditions. Explanations: 1—watersheds; 2—cone of depression range; 3—groundwater flow direction; 4—hydrogeological cross-section; 5—model boundary; 6— Water 2019, 11, 848 10 of 16 groundwater level in mining conditions [56].

N S Cone of depression range m n.p.m. Bełchatów Field

200

150

100

0 2 4km 0

5 7 89 1 2 34 6 Figure 6. Hydrogeological cross-section through Bełchatów Opencast Mine. Explanations: 1—ﬁne Figure 6. Hydrogeological cross-section through Bełchatów Opencast Mine. Explanations: 1—fine sands; 2—glacial tills; 3—limestones, marls and dolomites; 4—lignite; 5—stratigraphic boundary; sands; 2—glacial tills; 3—limestones, marls and dolomites; 4—lignite; 5—stratigraphic boundary; 6— 6—faults; 7—screened intervals of a weeks; 8—ground water level in pre-mining conditions; 9—ground Water faults;2019, 11 7, —screenedx FOR PEER intervalsREVIEW of a weeks; 8—ground water level in pre-mining conditions; 9—ground11 of 17 water level in mining conditions [57]. water level in mining conditions [57].

POLAND

WARSZAWA

WROCŁAW

KRAKÓW

14 A'

A M OSZYN-GROCHOWISKA Proposed Open Pit

Note?

LUBST覹

Note?

DRZEWCE

Warta

1 Warta 1 2 3 4 A' A 01234 5 km 5 6

Figure 7. Groundwater ﬂow modeling study for the proposed M ˛akoszyn-Grochowiska open pit in Figurethe Konin 7. Groundwater lignite basin for flow pre-mining modeling conditions. study for the Explanations: proposed M1—groundwaterąkoszyn-Grochowiska level in open pre-mining pit in theconditions; Konin lignite2—actual basin cone for pre-mi of depressionning conditions range; 3—model. Explanations: boundary; 1—groundwater4—groundwater level ﬂowin pre-mining direction; conditions;5—hydrogeological 2—actual cross cone section; of depression6—mining range; areas 3—model [59]. boundary; 4—groundwater flow direction; 5—hydrogeological cross section; 6—mining areas [59].

A M A' i M s M l i M M MMsMMn - MMMchMMisMa MMMMsiM M M s M M h a M n ą C g h

m. a. s.l. ć ć M . M l c M M J M i Ł l M M

l P . 100 N N J Ś 80 Q 60 Ng Q Q 40 Ng 20 CM 0 -20 -40

0 1, 0 2, 0 3, 0M m Ng CM

- 1 - 2 - 3 - 4 - 5 - 6 - 7

- 8 - 9 - 10

Figure 8. Hydrogeological cross-section through Mąkoszyn Grochowiska deposit. Explanations: 1— fine sands; 2—glacial tills; 3—silts; 4—clays; 5—marls and dolomites; 6—lignite; 7—Neogene- Paleogene-Mesozoic aquifer piezometric surface; 8—stratigraphic boundary; 9—screened intervals; 10—fault [59].

Water 2019, 11, x FOR PEER REVIEW 11 of 17

POLAND

WARSZAWA

WROCŁAW

KRAKÓW

14 A'

A M OSZYN-GROCHOWISKA Proposed Open Pit

Note?

LUBST覹

Note?

DRZEWCE

Warta

1 Warta 1 2 3 4 A' A 01234 5 km 5 6

Figure 7. Groundwater flow modeling study for the proposed Mąkoszyn-Grochowiska open pit in the Konin lignite basin for pre-mining conditions. Explanations: 1—groundwater level in pre-mining conditions; 2—actual cone of depression range; 3—model boundary; 4—groundwater flow direction; Water 2019, 11, 848 11 of 16 5—hydrogeological cross section; 6—mining areas [59].

A M A' i M s M l i M M MMsMMn - MMMchMMisMa MMMMsiM M M s M M h a M n ą C g h

m. a. s.l. ć ć M . M l c M M J M i Ł l M M

l P . 100 N N J Ś 80 Q 60 Ng Q Q 40 Ng 20 CM 0 -20 -40

0 1, 0 2, 0 3, 0M m Ng CM

- 1 - 2 - 3 - 4 - 5 - 6 - 7

- 8 - 9 - 10

FigureFigure 8. Hydrogeological 8. Hydrogeological cross-section cross-section through through Mąkoszyn M ˛akoszynGrochowiskaGrochowiska deposit. Explanations: deposit. Explanations: 1— fine1 —finesands; 2 sands;—glacial 2—glacialtills; 3—silts; tills; 4—clays;3—silts; 5—marls 4—clays; and dolomites;5—marls 6—lignite; and dolomites; 7—Neogene-6—lignite; Paleogene-Mesozoic7—Neogene-Paleogene-Mesozoic aquifer piezometric aquifer surface piezometric; 8—stratigraphic surface; 8boundary;—stratigraphic 9—screened boundary; intervals;9—screened 10—faultintervals; [59].10 —fault [59]. Water 2019, 11, x FOR PEER REVIEW 12 of 17 By the 1990s, the numerical models for the mining industry in Poland mainly solved the problem of theBy amount the 1990s, of mine the numerical water inflow models and for environmental the mining industry impact in Poland assessment mainly caused solved by the mine problem water drainage.of the amount The problem of mine of water post-mining inflow and excavations environmental flooding impact had not assessment been extensively caused by addressed mine water since theredrainage. were The not problem many such of post-mining reservoirs in excavations Poland. Presently, flooding had the not problem been extensively of filling post-mining addressed since voids withthere water were isnot more many extensively such reservoirs addressed in Poland. for the Presently, reasons the of estimatingproblem of thefilling costs post-mining of works voids related towith flooding water ligniteis more opencast extensively mines addressed (Figures for9 andthe re10asons). It isof foreseen estimating that the in costs the nextof works years, related the total to volumeflooding of lignite flooded opencast post-mining mines voids(Figures in Polish9, 10). ligniteIt is foreseen industry that will in the reach next more years, than the 3.5 total billion volume cubic metersof flooded (Table post-mining1). The most voids crucial in Polish elements lignite in modeling industry studieswill reach for more water than management 3.5 billion of cubic post-mining meters excavations(Table 1). The are: most The rate crucial of flooding, elements restoration in modelin ofg the studies groundwater for water table, management the impact of of post-mining the reservoirs onexcavations groundwater, are: andThe the rate changes of flooding, in the waterrestoratio qualityn of in the reservoir groundwater and aquifer. table, the impact of the reservoirs on groundwater, and the changes in the water quality in reservoir and aquifer. Table 1. Post-mining voids in the Polish lignite industry. Table 1. Post-mining voids in the Polish lignite industry. 1 Area Volume OpencastOpencast MinesDeposit Deposit Exploitation Number of Voids Area2 Volume3 Number of voids1 (km ) (mln m ) mines exploitation (km2) (mln m3) Adamów 1964–2018 8 10.6 252 Adamów 1964–2018 8 10.6 252 Konin 1946–2030 10 21.8 664 Konin 1946–2030 10 21.8 664 Bełchatów 1981–2038 2 32.5 2422 BełchatówTur ów1981–2038 1904–2040 2 1 32.5 17.0 1220 2422 Turów 1904–2040 1 17.0 1220 1 Some after a rehabilitation. 1 Some after a rehabilitation.

FigureFigure 9. 9.Current Current andand proposedproposed location of pit lakes lakes in in the the Adamow Adamow (left) (left) and and the the Konin Konin post-mining post-mining areaarea (right). (right). Explanations:Explanations: 1—natural lake; 22—river;—river; 33—post-mining—post-mining area; area; 4—pit4—pit lake lake [61]. [61 ].

Water 2019, 11, 848 12 of 16 Water 2019, 11, x FOR PEER REVIEW 13 of 17

BELCHAT覹

Kluki Ludwik體 KLUKI R GROCHOLICE Z . W ID A Wola Rusiecka W SZCZERC覹 K RUSIEC A Rusiec KASZEWICE

KA 覹 R AK Z R . Z. K R R A S 覹 R Z K . A W NowyJan體 E Parzniewice 桸 El.Be3chat體 IC A

LEKAWA ｊkawa Chabielice

CHABIELICE W3odzimierz

RKA EZIO RZ. J

R Z Kie3czyg3體 . W IE R

Z R A B Z I . K K C RZYO NIA RA

J A Str? a S覹 ZB. WINEK

E KA Z

R Rz1o nia STR訔 A

E 2110 KLESZCZ覹

Z 覹 LEKINSKO R SIEMKOWICE 覹 BELCHAT SZCZERC Kamiensk R Z. N IECI PIT LAKE PIT LAKE ECZ KAMIENSK ?obnica Lekinsko A ORK Siemkowice ZI E PO . J T RZ . K AM SULMIERZYCE IO S3ostowice NK Sulmierzyce GOMUNICE A 2100

R Krzywanice Z . Wiewi髍體 W ID A W 2090 K A

A PAJ蔆 ZNO IC Pajeczno-PI C E R K Debowiec . DOBRYSZYCE Z R 2080

DZIALOSZYN R Z . P IS I A Dobryszyce Krepa RZ . W AR 2070 TA 2050 Kietlin RADOMSKO Radomsko-Mi3aczki BRZE廚 ICA

Figure 10. Proposed location of pit lakes in the Belchatow post-mining area with the cone of depression Figure 10. Proposed location of pit lakes in the Belchatow post-mining area with the cone of reduction, in case of no additional recharge [57]. depression reduction, in case of no additional recharge [57]. 6. Limitations of Modeling Methods 6. Limitations of Modeling Methods Numerical models used in mining hydrogeology, especially in opencast mining, have certain limitationsNumerical and uncertainties models used related in mining to the hydrogeology, assumed stresses especially and parameters in opencast in the mining, model. Theyhave includecertain shortcomingslimitations and and uncertainties variability inrelated the boundary to the assume conditions,d stresses hydrogeological and parameters and in spatial the model. parameters, They groundwaterinclude shortcomings recharge, and mining variability schedules, in andthe others.boundary A regionalconditions, model hydrogeological to analyze the and effect spatial of an openparameters, pit on groundwatergroundwater levels recharge, often mining requires schedules, a lot more and discretization others. A toregional analyze model the slope to analyze stability the in theeffect same of an pit open [62]. pit The on finer groundwater nodal spacing levels is requiredoften requires to define a lot highly more curveddiscretization surfaces. to analyze It enables the a simulationslope stability of a in greater the same accuracy pit [62]. the declineThe finer in groundwaternodal spacing level is required and groundwater to define levelhighly itself curved near thesurfaces. slopes, It whichenables results a simulation in changes of of a geotechnicalgreater accuracy conditions the decline of slopes in ingroundwater time and space level as and the minegroundwater develops. level itself near the slopes, which results in changes of geotechnical conditions of slopes in timeThe and proper space identification as the mine develops. of hydraulic parameters, the nature of faults and fissures, as well as karsticThe formation proper identification phenomena, isof decisive hydraulic for parameters, the reliability the of nature modeling of faults studies. and For fissures, example, as naturalwell as hydraulickarstic formation barriers, phenomena, such as faults is decisive filled with for the semi-permeable reliability of modeling material studies. or rock formationsFor example, with natural low hydraulic conductivitybarriers, such compared as faults filled to adjacent with semi-permeable rock layers, can material decrease or the rock mine formations water inflow with andlow reducehydraulic the conductivity negative environmental compared to impactadjacent of rock dewatering. layers, can However, decrease whilethe mine the water deposit inflow is under and dewatering,reduce the negative new information environmental based onimpact the actual of de impactwatering. of theHowever, mine drainage while the on thedeposit groundwater is under anddewatering, surface new water information may result based in the on necessity the actual of im adjustingpact of the hydraulic mine draina parametersge on the of groundwater aquifers and semi-permeableand surface water layers may inresult the model.in the necessity Moreover, of oneadjusting should hydraulic be aware parameters that under of the aquifers mine dewatering and semi- hydraulicpermeable parameters layers in the may model. change Moreover, in time due one to should the activation be aware of a that new under groundwater the mine flow dewatering pathway. hydraulicDespite parameters the fundamental may change importance in time due of recharge to the activation in the water of a balance,new groundwater it is an element flow pathway. with the highestDespite uncertainty, the fundamental since its valuesimportance primarily of recharge depend in on the the water accuracy balance, assumed it is an whenelement calculating with the otherhighest elements uncertainty, of the since water its balance values equation.primarily Thoughdepend iton is the one a ofccuracy the most assumed important when components calculating inother groundwater elements of studies, the water recharge balance is equation. also one of Though the least it is understood, one of the most largely important because components recharge rates in varygroundwater widely in studies, space and recharge time, andis also rates one are of ditheffi cultleast to understood, directly measure largely [38 because]. Changes recharge in the rates recharge vary maywidely be thein space consequence and time, of manyand rates factors are whichdifficult were to notdirectly accurately measure estimated [38]. Changes or identified in the inside recharge and outsidemay be thethe miningconsequence area. These of many may factors include: which Changes were of not recharge accurately caused estimate by groundwaterd or identified fluctuation inside and outside the mining area. These may include: Changes of recharge caused by groundwater fluctuation and land management, stream-aquifer interaction, unidentified leakage of surface waters

Water 2019, 11, 848 13 of 16 and land management, stream-aquifer interaction, unidentified leakage of surface waters into aquifers, water losses from the water supply and sewage disposal system, and industrial waters recharging the aquifers. Moreover, mining and power generation activity, including the construction of large open pits and the overburden of dumping areas, as well as the existence of powerful heat and steam emitters at nearby power plants, may lead to local climate variations, influencing recharge from precipitation. Assuming the amount of aquifer recharge is based on average multiannual precipitation does not account for the cyclic deviations and trends resulting from climate variability. For instance, low precipitation may lead to a fall of the groundwater table below a level referred to as the evapotranspiration extinction depth, which causes an increase in effective infiltration. On the other hand, a change in the difference between potential evaporation and the precipitation affects the rate at which post-mining reservoirs are filled with water. Several years after the modeling study is completed, a post audit is conducted. Its goal is to determine whether the prediction was correct, based on new field data. Most often, errors in model forecasts are not the result of imperfections in programs, but mistakes made by modelers. In general, a post-audit studies found that errors in model predictions were caused by errors in the conceptual model of the hydrogeological system [28] and a failure to use appropriate values for assumed future stresses such as recharge and pumping rates [63,64]. In a modeling study for the simulation of opencast mine activity, the different boundary conditions which are not compatible with the actual conditions are the most serious problem. Due to changes in the schedules of mining activities or rehabilitation of post-mining excavations, the input data used in the model should be verified and updated with reference to the most recent recommendations and assumptions.

7. Conclusions Even though groundwater ﬂow models used in mining hydrogeology have numerous limitations related to the uncertainty of the parameters and boundary conditions, they still provide the most comprehensive information concerning the mine dewatering system and its environmental impact at the time when they are developed. However, they will always require periodical veriﬁcation based on new information on the actual response of the aquifer system to the mine drainage and the actual climate conditions, as well as up-to-date schedules of deposit extraction and mine closure. Numerical modeling used in mining hydrogeology needs a lot of experience, not only in hydrogeology but also in mining operations and other activities realted to mining industry in the neighboring areas.

Funding: This research received no external funding. Conﬂicts of Interest: The author declares no conﬂict of interest.

References

1. Szczepiński,J. The significance of ground water flow modeling study for simulation of open cast mine dewatering and assessing the environmental impact of drainage. In Proceedings Internatnal Mine Water Association Symposium, Leipzig, Germany, 11–15 July 2016; Drebenstedt, C., Michael, P., Eds.; TU Bergakademie Freiberg: Freiberg, Germany, 2016; pp. 508–511. 2. Papadopulos, I.S.; Cooper, H.H. Drawdown in a well of large diameter. Water Resour. Res. 1967, 3, 241–244. [CrossRef] 3. Marinelli, F.; Niccoli, W.L. Simple analytical equations for estimating ground water inflow to a mine pit. Groundwater 2000, 38, 311–314. [CrossRef] 4. Hanna, T.M.; Elfadil, A.A.; Atkinson, L.C. Use of analytical solution for preliminary estimates of ground water inflow to a pit. Min. Eng. 1994, 46, 149–152. 5. Powers, J.P.; Corwin, A.B.; Schmall, P.C.; Kaeck, W.E. Construction Dewatering and Groundwater Control. New Methods and Applications, 3rd ed.; John Wiley and Sons: Hoboken, NJ, USA, 2007; pp. 1–638. ISBN 978-0-471-47943-7. 6. Schwartz, F.W.; Crowe, A.S. Simulation of changes in ground-water levels associated with strip mining. Geol. Soc. Am. Bull. 1985, 96, 253–262. [CrossRef] Water 2019, 11, 848 14 of 16

7. Atkinson, L.C.; Keeping, P.G.; Wright, J.C.; Liu, H. The challenges of dewatering at the victor diamond mine in Northern Ontario, Canada. Mine Water Environ. 2010, 29, 99–107. [CrossRef] 8. Charou, E.; Stefouli, M.; Dimitrakopoulos, E.; Vasiliou, E.; Mavrantza, O.D. Using remote sensing to assess impact of mining activities on land and water resources. Mine Water Environ. 2010, 29, 45–52. [CrossRef] 9. Rapantova, N.; Grmela, A.; Vojtek, D.; Halir, J.; Bedrich, M. Ground water flow modeling applications in mining hydrogeology. Mine Water Environ. 2007, 26, 264–270. [CrossRef] 10. Banks, D. A variable volume, head dependent mine water filling model. Ground Water 2001, 39, 362–365. [CrossRef] 11. Bair, E.S.; Parizek, R.R. Numerical simulation of potentiometric surface changes caused by a proposed open-pit anthracite mine. Ground Water 1981, 19, 190–200. [CrossRef] 12. Bair, E.S.; O’Donnell, T.P. Uses of numerical modeling in the design and licensing of dewatering and depressurizing systems. Ground Water 1983, 21, 411–420. [CrossRef] 13. Castro, M.; Moore, J.N. Pit lakes, their characteristics and the potential for their remediation. Environ. Geol. 2000, 39, 1254–1260. [CrossRef] 14. Keqiang, H.; Dong, G.; Xianwei, W. Mechanism of the water invasion of Gaoyang iron mine, China and its impacts on the mine groundwater environment. Environ. Geol. 2006, 49, 1163–1172. [CrossRef] 15. Vukeliˇc,Ž.; Dervariˇc,E.; Šporin, J.; Vižintin, G. The development of dewatering predictions of the Velenje coalmine. Energies 2016, 9, 702. [CrossRef] 16. Younger, P.L.; Banwart, S.A.; Hedin, R.S. Mine Water Hydrology, Pollution, Remediation, 1st ed.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 2002; pp. 1–442. ISBN 1-4020-0137-1. 17. Wolkersdorfer, C. Water Management at Abandoned Underground Flooded Mines. Fundamentals, Tracer Tests. Modeling, Water Treatment, 1st ed.; Springer: Berlin, Germany, 2008; pp. 1–465. ISBN 978-3-540-77330-6. 18. Castendyk, D.N.; Eary, L.E. Mine Pit Lakes: Characteristics, Predictive Modeling and Sustainability (Management Technologies for Metal Mining Influenced Water, 1st ed.; Society for Mining, Metallurgy, and Exploration: Littleton, CO, USA, 2009; Volume 3, pp. 1–304. ISBN 978-0-87335-305-2. 19. Geller, W.; Schultze, M.; Kleinman, R.; Wolkersdorfer, C. Acidic Pit Lakes. The Legacy of Coal and Metal Surface Mine, 1st ed.; Springer: Berlin, Germany, 2013; pp. 1–525. ISBN 978-3-642-29383-2. 20. Remson, I.; Appel, C.A.; Webster, R.A. Groundwater models solved by digital computers. J. Hydraul. Div. Am. Soc. Civ. Eng. 1965, 91, 133–147. 21. Pinder, G.F.; Bredehoeft, J.D. Application of the digital computer for aquifer evaluation. Water Resour. Res. 1968, 4, 1069–1093. [CrossRef] 22. Prickett, T.A.; Longuist, C.G. Selected digital computer techniques for groundwater resources evaluation. Ill. State Water Surv. 1971, 55, 1–67. 23. Trescott, P.C.; Larson, S.P. Documentation of Finite-Difference Model for Simulation of Three—Dimensional Ground-Water Flow; Open File Report 75–438; U.S. Department of the Interior: Washington, DC, USA, 1976; pp. 1–21. 24. Bredehoeft, J. Modeling groundwater flow—The beginnings. Groundwater 2012, 50, 325–329. [CrossRef] 25. McDonald, M.G.; Harbaugh, A.W. A Modular Three-Dimensional Finite Difference Ground-Water Flow Model; Open File Report 83–875; U.S. Department of the Interior: Washington, DC, USA, 1984; pp. 1–528. 26. McDonald, M.G.; Harbaugh, A.W. The history of modflow. Groundwater 2003, 41, 280–283. [CrossRef] 27. Diersch, H.J.G. FEFLOW Finite Element Modeling of Flow, Mass and Heat Transport in Porous and Fractured Media, 1st ed.; Springer: Berlin, Germany, 2014; pp. 1–808. ISBN 978-3-642-38738-8. 28. Anderson, M.P.; Woessner, W.W. Applied Groundwater Modeling, Groundwater Flow and Adjective Transport, 1st ed.; Academic Press Inc.: San Diego, CA, USA, 1992; pp. 1–381. ISBN 9781483999785. 29. Spitz, K.; Moreno, J. A Practical Guide to Groundwater and Solute Transport Modeling, 1st ed.; John Wiley and Sons, Inc.: New York, NY, USA, 1996; pp. 1–461. ISBN 0-471-13687-5. 30. Ritchey, J.D.; Rumbaugh, J.O. Subsurface Fluid Flow (Ground-Water and Vadose Zone) Modeling, 1st ed.; ASTM Special Technical Publication 1288; ASTM: West Conshohocken, PA, USA, 1996; pp. 1–423. ISBN 0-8031-2021-4. 31. D ˛abrowski, S.; Kapu´sciński, J.; Nowicki, K.; Przybyłek, J.; Szczepański, A. Metodyka Modelowania Matematycznego w Badaniach i Obliczeniach Hydrogeologicznych, 1st ed.; Bogucki Wydawnictwo Naukowe: Poznań,Poland, 2011; pp. 1–364. ISBN 978-83-62662-41-8. Water 2019, 11, 848 15 of 16

32. Martinez, C.; Ugorets, V. Use of numerical modeling for mine dewatering assessment. In Proceedings of the 2nd International Congress on Water Management in the Mining Industry, Santiago, Chile, 9–11 June 2010; pp. 317–326. 33. Szczepiński, J. Modelowanie Numeryczne w Badaniach Hydrogeologicznych dla Oceny Wpływu Kopalń Odkrywkowych na Srodowisko´ Wodne, 1st ed.; Wydz. Geoin˙zynierii, Górnictwa i Geologii Politechniki Wrocławskiej: Wrocław, Poland, 2013; pp. 1–200. ISBN 978-83-978-83-937788-0-5. 34. Franke, O.L.; Reilly, T.E.; Bennett, G.D. Definition of Boundary and Initial Conditions in the Analysis of Saturated Ground-Water Flow Systems—An Introduction; U.S. Geological Survey: Reston, VA, USA, 1987; Chapter B5; pp. 1–15. 35. Franke, O.L.; Reilly, T.E. The Effects of Boundary Conditions on the Steady-State Response of Three Hypothetical Ground-Water Systems—Result and Implications of Numerical Experiments; Paper 2315; U.S. Geological Survey: Reston, VA, USA, 1987; pp. 1–19. 36. Reilly, T. System and Boundary Conceptualization in Ground-Water Flow Simulation; U.S. Geological Survey: Reston, VA, USA, 2001; Chapter B8; pp. 1–26. 37. Reilly, T.E.; Harbaugh, A.W. Guidelines for Evaluating Ground-Water Flow Models. U.S. Geological Survey Scientific Investigations Report 2004–5038; 2004; pp. 1–30. Available online: https://pubs.usgs.gov/sir/2004/ 5038/PDF/SIR20045038_ver1.01.pdf (accessed on 2 June 2018). 38. Healy, R. Estimating Groundwater Recharge, 1st ed.; Cambridge University Press: Cambridge, UK, 2010; pp. 1–256. ISBN 978-0-521-86396-4. 39. Rushton, K.R. Numerical and Conceptual Models for Recharge Estimation in Arid and Semi-Arid Zones; Simmers, I., Ed.; Reidel: Dordrecht, The Netherlands, 1988; pp. 223–238. 40. De Vries, J.J.; Simmers, I. Groundwater recharge, an overwiew of processes and challenges. Hydrogeol. J. 2002, 10, 5–17. [CrossRef] 41. Sawicki, J. Zmiany Naturalnej Infiltracji Opadów do Warstw Wodono´snychpod Wpływem Gł˛ebokiegoDrena˙zu Górniczego; Ofic. Wydaw. Politechniki Wrocławskiej: Wrocław, Poland, 2000; p. 174. ISBN 83-7085-475-3. 42. Sanford, W. Recharge and groundwater models. Hydrogeol. J. 2002, 10, 110–120. [CrossRef] 43. Harbaugh, A.W.; Banta, E.R.; Hill, M.C.; McDonald, M.G. MODFLOW-2000, The U.S. Geological Survey Modular Ground-Water Model—User Guide to Modularization Concepts and the Ground-Water Flow Process; U.S. Geological Survey: Reston, VA, USA, 2000; pp. 1–121. 44. Szczepiński,J. Predictive modeling of filling with water the Bełchatów lignite mine final excavations. In Proceedings of the 7th International Mine Water Congress, Katowice, Poland, 11–15 September 2000;Ró˙zkowski,A., Rogo˙z,M., Eds.; The IMWA and Uniwersytet Sl´ ˛aski:Katowice, Poland, 2000; pp. 331–341. 45. Anderson, M.P.; Munter, J.A. Seasonal reversals of groundwater flow around lakes and the relevance to stagnation points and lake budgets. Water Resour. Res. 1980, 17, 1139–1150. [CrossRef] 46. Morgan, D.S. Geohydrology and Numerical Model Analysis of Groundwater Flow in the Goose Lake Basin, Oregon and California; Report 00–4167; U.S. Geological Survey: Reston, VA, USA, 1988; pp. 1–101. 47. Anderson, M.P.; Hunt, R.J.; Krohelski, J.T.; Chung, K. Using high hydraulic conductivity nodes to simulate seepage lakes. Groundwater 2002, 40, 117–122. [CrossRef] 48. Hunt, R.J.; Krohelski, J.T. The application of the analytic element model to investigate groundwater-lake interactions at Pretty lake. Wis. J. Lake Reserv. Manag. 1996, 12, 487–495. [CrossRef] 49. Lee, T.M. Hydrogeologic controls on the groundwater interactions with an acid lake in karst terrain, Lake Barco, Florida. Water Resour. Res. 1996, 32, 831–844. [CrossRef] 50. Hunt, R.; Haitjema, H.M.; Krohelski, J.T.; Feinstein, D.T. Simulating ground water-lake interactions, approaches and insights. Groundwater 2003, 41, 227–237. [CrossRef] 51. Szczepiński,J. Forecasting of groundwater table recovery after closure of the bełchatów lignite open cast mine. In Proceedings of the 8th International Mine Water Congress, Belo Horizonte, Brazil, 24–28 April 2001; pp. 1–6. 52. Doherty, J. Improved calculations for dewatered cells in MODFLOW. Groundwater 2005, 39, 863–869. [CrossRef] 53. HydroGeoLogic, Inc. MODFLOW-SURF (Version 1.2). The HydroGeoLogic Enhanced Version of the U.S. Geological Modular Groundwater Flow Model. Code Documentation Report. 1996. 54. Szymanko, J. Zastosowanie modelowania matematycznego przy projektowaniu odwodnieńwykopów metra na przykładzie wykopu stacji A-10. Przegl ˛adGeol. 1987, 35, 375–380. Water 2019, 11, 848 16 of 16

55. Bocheńska,T.; Fiszer, J. Okre´slenieWspółczynnika Filtracji CechszytyńskichW˛eglanów w-1 na Drodze Modelowania Cyfrowego z Wykorzystaniem Metody Elementów SkończonychTechnika Poszukiwań. 1983; Volume 58, pp. 22–28. Available online: http://pph.psh.gov.pl/apex/pph.getfile?p_id=12228838331743 (accessed on 2 June 2018). 56. Szczepiński,J.; Libicki, J. Modeling of the Bełchatów open-pit impact on the groundwater environment. In Proceedings of the International Mine Water Symposium, Sevilla, Spain, 13–17 September 1999; pp. 69–77. 57. Szczepiński,J. Model odwadniania kopalni odkrywkowej w˛eglabrunatnego „Bełchatów”. In Metodyka Modelowania Matematycznego w Badaniach i Obliczeniach Hydrogeologicznych, 1st ed.; D ˛abrowski, S., Kapu´sciński,J., Nowicki, K., Przybyłek, J., Szczepański,A., Eds.; Bogucki Wydawnictwo Naukowe: Poznań,Poland, 2011; pp. 244–262. ISBN 978-83-62662-41-8. 58. Szczepiński,J. Modeling of the future “Piaski” open pit impact on the water environment. In Proceedings of the 10th Intern. Mine Water Congress, Karlove Vary, Czech Republic, 2–5 June 2008; Rapantowa, N., Hrkal, Z., Eds.; VSB-Technical University of Ostrava: Ostrava, Czech Republic, 2008; pp. 481–484. 59. Szczepiński,J.; Straburzyńska-Janiszewska,R. Prognoza zasi˛eguleja depresji dla odkrywki w˛eglabrunatnego M ˛akoszyc-Grochowiska KWB „Konin” S.A. Biuletyn Pañstwowego Instytutu Geologicznego 2011, 445, 671–684. 60. Libicki, J. Hydrogeological conditions of lignite basins in Poland and their changes cause by dewatering of open pits. In Proceedings of the International Mine Water Association Symposium, Katowice, Poland, 22 September 1987; pp. 525–535. 61. Szczepiński,J.; Fiszer, J.; Stachowicz, Z.; Szczepanik, P. Reclamation of Polish lignite open pits by flooding. Biuletyn Pañstwowego Instytutu Geologicznego 2010, 441, 167–174. 62. Martinez, C.; de la Hoz, K.; Pereira, C. Groundwater flow modeling applications in mining, scopes and limitations. In Proceedings of the 2nd International Congress on Water Management in the Mining Industry, Santiago, Chile, 9–11 June 2010; pp. 267–274. 63. Konikow, L.F. Predictive accuracy of ground-water model—Lessons from a Postaudit. Ground Water 1986, 24, 173–184. [CrossRef] 64. Andersen, P.F.; Lu, S. A post audit of a model-designed ground water extraction system. Ground Water 2003, 41, 212–218. [CrossRef][PubMed]

© 2019 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).